1TO PLOEA INDICA. [Meniqperwacea. spherical cotyledons have the ordinary position. The seed has therefore an elon- gated cylindrical shape, and is curved like a horse-shoe or siphon, the hilum occn- pying the "bottom of the concavity. The putamen then forms a bony sheath, which closely invests the seed; hut the concavity of the horse-shoe, along which the nutri- tive vessels run from the base of the fruit, is filled up by one or more bony plates, variously perforated, and sometimes enclosing empty spaces. The albumen of Menispermacea varies as much as the form of the embryo. It is generally only present in small quantity, and in the tribe 'Pachygonece it is entirely wanting. Most commonly it is fleshy and homogeneous; but in several genera not otherwise very closely allied, namely in Tinospora, Abuta, and Tiliacora, it is very oily, and ruminated by means of transverse membranous plates. In Jnamirta it contains small granular masses of a different texture from the greater portion; and finally, in Coscinium it is irregularly ruminated by cellular plates springing from the hilum, the nature of which has not been accurately determined. The close relationship of Menispermacea to the great class of Apocarpous Thala- miflora, in which they are generally placed, may be considered well established; as the ingenious arguments by which Dr. Lindley attempts to maintain bis opinion that they are more nearly related to apetalous orders have been well answered by M. Decaisne; who has shown, as we think, successMLy/t/Ut neither the structure'of the wood (to which we shall advert more particularly further on) nor the unisexual flowers, are to be relied on as indications of affinity. To all the Orders of this great class, Menispermacea present more or less affinity by means of aberrant species, though the typical forms collectively possess,such a peculiar habit as to make them a very natural family. With Anonacece they are con- nected by means of the genera with ruminated albumen; with Myristicacea through Coscinium; with Lardizalalacea and Berberidece through Burasaia; to Uanuncu- lacets they are only allied through Berberide^^ and to M&gnoli&cea through Schi- zandracear. Dilleniacea are the most distant, lying at the opposite extremity of the class, so as to form a passage to a very different series of Orders. Menispermaeea agree with Bertieridea in the structure and number of the parts of the perianth, in'the usually definite stamens, in the solitary ovarmm of Cissam- pelidea, and in the comparatively large embryo, differing, however, in many impor- tant points.' From Lardizalalacea, which they approach very closely in the number of parts and in the dioecious flowers, Menispermacete are readily distinguished by their solitary ovules. The remarkable position of the indefinite ovules of all the genera of IMF- dizabalacea except Decaitnea, the anatropous seeds, and the minute embryo, are other important distinctions. The compound leaves of Lardizabalacece exist in Buraaaia, which was by Deeaisne doubtfully referred to that Order, but which Mr. Miers places in Menispermacece. Its structure appears to us to be quite interme- diate between the two; but though the anatropous ovules are anomalous in Menisper- macfa, the seed has, according to Thouars, the divaricating cotyledons of the tribe Tinosporece. The genus Lardizabala has amphitropous seeds, excavated on one side, so as to resemble those of Tinospora, but their minute embryo is not Menispermeous. Anonacea, which in general are BO very distinct in habit and characters, are yet immediately connected with Meni*permace<s by those genera which have definite stamens, as well as by the remarkable occurrence of ruminated albumen in several genera of Menispermaeea. The abnormal genus fycnarrhena approaches in habit to such aberrant Anonacecs as Stelechocarpus and Guatteria pallida, Schizandracea form the link which connects Henispermacete with Magnoliacea; bat the relationship is not very near, except by means of Sabia, which is very closely allied to both Orders; and by the amphitropous ovules of both. Notwithstanding the close relationship which is now fully 'established as existing Natural Orders, and are not of universal occurrence in them. Mr. Micrs' condyle we shall call processus intemus putaminis, and we shall designate it as condi/liformis, when (as in Tiliacora) it resembles a condyle in form.